Dec 7th, 2009, 04:57 AM
http remote invoker with rich client and temporary connection
Hi spring community,
we already do use http remote invoker to access services running on an application server, actullay a jetty, from an eclipse rcp application. We export the services imported via http as osgi services. That works fine. In production environments nearly every system has to deal with unreliable network connections between server and client, temporarily shut down servers and the like. So far I have not been able to find a "best practice" how to deal with these issues and provide a good user experience. I definitely want the user to know whether I do not know a service of a certain type or if I know it but I am not able to reach it. And of course I do want to "monitor" if a remote service is available or not. Two possibilities that do work but may not be the best practice, if there is one, are 1.) call a (ping) method on the remote service and catch remote exception to determine the state or 2.) Get head from http url and evaluate (prodcues exception on server side spring proxy). The reachability of the service computed by one of the methods described above is exported along with the service and we are done. Nearly. The UI part has to be done but that is not spring and not remoting. Does anybody have an idea, link, document or something dealing with monitoring the reachibility of remote services ?
Tags for this Thread